GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 1.1 GNGTS 2023 between these events can be computed with high accuracy. The location method incorporates ordinary absolute travel-time measurements and/or cross correlation P and S wave differential travel-time measurements. Even though HypoDD can be applied to standard catalogue data, the analysed dataset is better exploited, and produces more accurate results, when a database of digital waveforms is available for the application of the cross-correlation technique. In our study, we inverted both travel-time data and cross-correlations, when the relative seismograms were available. Intensity inversion by Kinematic Function KF Over a period of more than 20 years, the Kinematic Function KF is used to retrieve source information by inverting the macroseismic pattern of historical earthquakes. Already descripted in many papers (Sirovich 1996: Sirovich and Pettenati 2001), KF is a simple model composed by radiation pattern of S-waves, the geometrical spreading and the directivity (see formula [1] in Sirovich and Pettenati 2001). The body-wave radiation is produced by a linear source in an elastic medium in the distance range 5-100 km from the nucleation point. The inversion is driven by the simple step-by-step version (Sirovich and Pettenati 2001) or the Genetic Algorithm with Niching version - NGA (Sirovich and Pettenati 2004; Pettenati and Sirovich, 2007). The KF has been applied to two historical earthquakes of the Lunigiana-Garfagnana, namely the 1920 and the 1837 events. The 7 September 1920 Mw 6.5 earthquake (Eva et al., 2022) was inverted by the NGA version of KF. The used dataset with about 500 intensity points has been extracted from the DBMI15 database (Locati et al. 2016). The result is a dip-slip source, interpreted as a normal fault, with strike SE dipping SW, located in and aligned with the Lunigiana graben system. The solution is likely associated with the SW dipping Compione-Comano fault in its southern termination at a depth of 14 km (±4) (Eva et al., 2022). The 11 April 1837 Mw 5.8 Lunigiana earthquake was inverted by a step-by-step version. The data used, again taken from the DBMI15 database (Locati et al. 2022), is composed of 54 intensities (Fig. 2), with only one X degree, an epicentral cluster of VIII and VII intensities and some spread sites with lower intensity. The inversion provided two solutions with similar fit, which is the parameter used to estimate the robustness of the result. One solution is a dip-slip with strike for NNE-SSW, M=6.1 and fit 20. The second solution has fit 21 and is a dip-slip trending NE-SW, M=5.8, the same of the catalogue CPTI15 (Rovida et al., 2016). In order to make a validation test, we got rid of all intensities far from the epicentral area, inverting the 46 sites close to the epicentre. The inversion confirms the solution with strike NE-SW, of course with a lower fit value (9). We then choose the NE-SW as the best solution also because, in the case of 46 sites, the χ 2 (fit/number of data) is the lowest.